ABR Certifying Exams In Diagnostic Radiology Early Results Future Considerations 1
Exam Goals: Core Overarching goal: to protect the public by determining that individual candidates have attained competence in basic diagnostic radiology Specific goal: to create examinations that are relevant to current radiologic practice Specific goal: to make a reliable pass/fail decision about candidate performance in each category For this high-stakes exam, reliability requires ~60 questions/category Emphasis on unique importance of Physics content = 136 questions 2
Core Exam in Diagnostic Radiology Now three administrations: October, 2013 June, 2014 October, 2014 Covers all of diagnostic radiology Comprehensive, categorical exam Candidates must pass all categories 3
What Are These Categories? Organ systems MSK, Thoracic, GI, Urinary, Neuro, Pediatrics, Cardiac, Reproductive/Endocrine, Mammography, Vascular Modalities CT, MRI, RF/Fluoro, Nucs, Interventional, Ultrasound Fundamentals Physics, Safety 4
Breast Cardiac GI MSK Neuro Peds Thorax Repro / Endo Urinary Vascular Q# CT 85 IR 60 MR 89 NM/Molecular 81 Rad/Fluoro 73 US 79 Physics 136 Safety 71 Q# 60 60 60 60 60 60 60 60 60 60 5
Core Exam Statistics 1 st Administration: 1206 Candidates 980 (79%) in Chicago 226 (21%) in Tucson 2 nd Administration: 1413 Candidates Same distribution between cities Two different examination forms created from identical blueprints 6
How Did It Go? Survey sent to all candidates after 2013 exam 647 (54%) responded Similar survey after 2014 exam 815 (58%) responded 7
Getting Started Renaissance Hotel generally worked Relatively inexpensive (for Chicago) per night (new: negotiated room rate includes breakfast and internet) Worked well as a staging location Ran out of blocked rooms (partially fixed in 2014) Registration Palm vein scanner worked well in Tucson Worked poorly for re-registration in Chicago Currently re-evaluating now using photo ID method 8
Exam Center Generally positive (both centers) Chicago two frequent comments Floor shakes Cubicle dividers not high enough Snacks BYO Time added to exam to provide for breaks 9
Exam Delivery Enough time to finish (657 scoreable units) Print/image layout good Some images too small, became pixelated Sonography, especially cine loops Cardiac Chest radiographs (especially laterals) 11
Most Frequent Candidate Concerns 1) Why Chicago/Tucson? 2) Cine vs. scrolling Cine loops difficult to control Will probably be used only on cardiac and US 3) Drop-down menu malfunction Software glitch not discovered in testing ABR rescored exams with/without those items 12
Core Exam: Content Combines old written and oral content ~40% require fact retrieval ~60% require higher-order analysis Development of differential diagnosis Recognition of management methods Majority of items linked with an image (effort made to emulate workstation to some degree) What is the most likely diagnosis? Most physics items practical What is this artifact, and how would you fix it? How would you reduce dose? Additional physics content: radionuclide safety 13
Example: Extended Matching For each patient whose clinical and imaging information is shown, select the most appropriate diagnosis from the list below. Each option may be used once, more than once, or not at all. A. Focal nodular hyperplasia B. Liver cell adenoma C. Cavernous hemangioma D. Inflammatory pseudotumor E. Pyogenic abscess F. Fungal abscess G. Nodular focal fat H. Biliary cystadenoma I. Lymphoma J. Solitary metastasis K. Intrahepatic cholangiocarcinoma L. Hepatocellular carcinoma M. Fibrolamellar carcinoma 14
1: 35 year old woman who underwent sonographic evaluation for mild abdominal discomfort. She was referred for MR imaging to characterize a solitary liver mass seen on that ultrasound. Images are obtained 30 seconds (A), 70 seconds (B) and 1 hour (C) after administration of gadobenate intravenously. A B C A. Focal nodular hyperplasia B. Liver cell adenoma C. Cavernous hemangioma D. Inflammatory pseudotumor E. Pyogenic abscess F. Fungal abscess G. Nodular focal fat H. Biliary cystadenoma I. Lymphoma J. Solitary metastasis K. Intrahepatic cholangiocarcinoma L. Hepatocellular carcinoma M. Fibrolamellar carcinoma Key = A 15
2. 60 year old man with abdominal pain. Imaging performed elsewhere showed a liver mass, and he is referred for CT imaging to characterize it. Images are obtained before (A), 20 seconds after (B), 50 seconds after (C), and 5 minutes after (D) intravenous administration of iodinated contrast material. A. Focal nodular hyperplasia B. Liver cell adenoma C. Cavernous hemangioma D. Inflammatory pseudotumor E. Pyogenic abscess F. Fungal abscess G. Nodular focal fat H. Biliary cystadenoma I. Lymphoma J. Solitary metastasis K. Intrahepatic cholangiocarcinoma L. Hepatocellular carcinoma M. Fibrolamellar carcinoma Key = L 16
a b c d e f g h A coronal CT image obtained in a patient with ascites is shown. Label the following structures: a) Left subphrenic space b) Lesser sac, inferior recess c) Lesser sac, superior recess d) Transverse mesocolon e) Gastrohepatic ligament f) Morison s pouch g) Left paracolic gutter h) Root of intestinal mesentery 17
How Did They Do? Number of problem items less than usual Historically, on written = 10% On both Core exams = 4% Pass rate very close to expected (comparing to previous Written/Oral first-time takers) 2013: 12% fail, 1% condition 2014: 8% fail, 1% condition Condition rate for Physics comparable to rates for Physics written 18
Feedback: Program 19
Feedback: Individual 20
21
Core Examination: Impact on Training Candidates must be exposed to all basic diagnostic radiology by end of third year Core anxiety and review sessions will occur midway through third year Content of review sessions will change(?) 22
Core Exam: Review Sessions Physics content Chief focus on practical applications Physics resource: rsna.org/education/physics.cfm Not intended to be sole resource Whatever you did seems to have worked! Diagnostic content 40% fact retrieval even though image-rich, reasonable to prepare as for past written exam 60% resembles oral boards standard board review Resource: Core study guide on theabr.org Annotated study guide (flags topics that have been made into questions) Blueprints for all categories Practice Exam 23
24
Core Exam: Potential Benefit In some programs (ours is one) 4 th year residents can, if they choose, take clinical rotations to prepare themselves for subspecialized practice Majority of our residents have taken advantage (three 3-month selectives ) 25
Core Exam: Future Plans Distributed Exam (local test center) Initial aim was 2018 Difficulties with that timeline (Screen resolution) Ambient light control Modular exam design Image-rich = very large! 26
Core Exam: Assessment Early results suggest that the Core is a discriminating, reliable exam Not so easy to determine whether it tests what we want it to test (i.e., protects the public) Actively seek feedback from stakeholders re outcome of new examination process 27
Core Exam: Item Writing Ideas for future testing Structured reporting (avoids cuing the answer) Simulation (especially in interactional/interventional categories) Computer adaptive testing Need more volunteers! 28
Certifying Exam in Diagnostic Radiology Will be first administered in October, 2015 Contains five parts (but questions will appear in random order) Non interpretive skills Essentials Clinical Practice Areas (3 self-selected CPAs) Things every physician should know Things every diagnostic radiologist should know Things this specific radiologist should know 29
Clinical Practice Areas (CPA) Twelve categories Organ system: MSK, Cardiac, Thoracic, GI, Urinary, Neuro, Peds, Breast Technology: US, VIR, Nuclear Radiology General Candidates can choose any combination of CPAs Items will vary in both difficulty and scope Level 1: fundamental Level 2: advanced Candidates selecting a CPA more than once will receive a higher proportion of level 2 items 30
Certifying Exam: Content (CPA) Emulates clinical practice Will include normals and variants Will include important findings outside chosen area Appropriateness, clinical vignettes, management decisions Item types Familiar: MCQs, extended matching Unfamiliar*: Structured reporting, simulation *Examples posted on ABR Website >1 year before use 31
Certifying Exam: Goals To confirm candidate has acquired and maintained necessary skills to practice independently Dual role First exam of practice-based learning (like MOC) Final ABR Certification Exam 32
Certifying Exam: Nuts and Bolts ~310 items (60/module + RISE content) 5-hour exam (one half-day) Exam center(s): Chicago, Tucson Exams offered October/March or April Scoring: Criterion-referenced exam (not on the curve ) Experience with MOC suggests high pass rate 33
Certifying Exam: Resources Non-interpretive skills and Essentials Study guides posted on theabr.org ~100 page document, posted January, 2013 Clinical practice modules Study guides at http://www.theabr.org/ic-drcertifying-exam CME, SAMs Practice-based learning Society-produced educational modules 34
Questions? balfed@mir.wustl.edu or balfe@theabr.org 35
Scoring the Categories Analogy to old Oral exam: 69 = below standard, but may be raised if global performance satisfactory 68 = well below standard, performance too poor to be raised Category passing standard determined by Angoff process (mean Angoff score for all questions in the category) 36
Score Category Scoring Process Angoff-defined passing standard Scores falling in this range were raised if test performance above passing standard 68 too low to be raised Category GI Scores falling in this range were not raised: candidate conditions GI 37
Score Category Scoring Process: Physics Angoff-defined passing standard Scores in this range were not raised: candidate conditions Physics Category Physics Physics had enough questions (136) that a pass/fail decision could be made using the Angoff-derived passing standard (reliability > 0.8) 38
2013 Core Results 146 (12%) did not meet the passing standard Of the (88%) who met the overall standard, 16 (1.5%) conditioned physics No candidate who met the passing standard fell below the 68 cutoff in any other category 83 candidates who met the passing standard were below the Angoff in one or more categories, but above the 68 cutoff in that category the scoring model raised all of these virtual 69s All candidates who fell below the 68 cutoff in any category did not meet the overall exam passing standard 39